Before After Image Editing Prompts: 7 Copy-Ready GPT Image 2 Workflows

The best before after image editing prompts describe the original image, the exact change, the details that must stay fixed, and the pass/fail test for the result. Use the prompts below as copy-ready starting points for GPT Image 2 when you want a visible transformation without losing identity, composition, product accuracy, or text clarity.

Most weak edits fail because the prompt only says make it better, turn this into aesthetic, or edit like this. A stronger prompt treats the edit like a production brief: before state, after state, locked elements, edit strength, and output use.

Use these examples in the GPT Image 2 prompt library, then run or adapt them in the GPT Image 2 workspace.

The before/after prompt formula

Use this structure whenever you are editing an existing image:

Edit the provided image.

Before state:
[Describe what is currently visible: subject, pose, room, product, lighting, background, text, camera angle.]

After goal:
[Describe exactly what should change.]

Keep fixed:
[Identity, facial features, product shape, logo, camera angle, layout, hand position, clothing, text, proportions.]

Edit strength:
[Subtle / medium / dramatic], but preserve natural realism and avoid changing the core subject.

Output:
[Aspect ratio, crop, final use, quality bar.]

Quality check:
The final image should clearly show the change from before to after while keeping the locked details consistent.

The Keep fixed line matters most. It tells GPT Image 2 which parts are identity-critical and should not be reimagined.

1. Before/after storyboard prompt

Use a storyboard when you need several edit stages instead of one final frame. This is useful for creative direction, renovation planning, beauty retouching, product cleanup, or client approval decks.

Edit the provided image into a clean before/after storyboard.

Create a 3x2 storyboard grid:
Frame 1: original before state, kept close to the source image.
Frame 2: mark the main problems with subtle visual callouts.
Frame 3: intermediate edit with improved lighting and cleaner background.
Frame 4: final after edit with polished color, balanced contrast, and corrected distractions.
Frame 5: close-up detail showing the most important preserved feature.
Frame 6: final comparison panel labeled BEFORE and AFTER.

Keep fixed:
Do not change the subject identity, pose, product shape, logo, camera angle, or core composition.

Style:
Professional creative review board, clean editorial spacing, neutral labels, realistic image treatment.

Quality check:
The viewer should understand what changed, what stayed fixed, and why the final version is better.

Why it works: the prompt asks for a comparison system, not just a prettier image. It forces the model to preserve the original frame while showing an edit path.

2. Portrait glow-up prompt

This is for a realistic before/after portrait edit, not a new person.

Edit the provided portrait photo.

After goal:
Create a polished but natural portrait glow-up. Improve skin texture gently, reduce harsh shadows, brighten the eyes slightly, clean small distractions, and add soft editorial lighting.

Keep fixed:
Keep the person's face shape, age, hairstyle, expression, clothing, pose, camera angle, and background structure. Do not change identity. Do not over-smooth skin.

Edit strength:
Medium polish with realistic texture.

Output:
Vertical portrait crop, high-resolution, natural color, suitable for a profile photo or editorial headshot.

Quality check:
The after image should look like the same photo after professional retouching, not a different person.

If the result looks too artificial, add: preserve pores, natural asymmetry, and original facial proportions.

3. Background replacement prompt

Background edits fail when the new scene ignores the original lighting. Anchor the replacement to the source image first.

Edit the provided image by replacing only the background.

Before state:
The subject is already correctly framed. The current background is distracting and should be replaced.

After goal:
Replace the background with a clean modern studio environment with soft depth, subtle shadows, and a calm premium color palette. Match the lighting direction and lens feel of the original subject.

Keep fixed:
Keep the subject, pose, edges, clothing, product details, facial features, and foreground shadows. Do not crop the subject differently.

Edit strength:
Dramatic background change, subtle subject integration.

Output:
Realistic final image, no cutout halo, no floating subject, no mismatched perspective.

Quality check:
The background should look newly designed, but the subject should still look photographed in the same scene.

For ecommerce, replace modern studio environment with the actual surface, wall color, props, and brand palette.

4. Product before/after cleanup prompt

Use this when you have a product photo with dust, glare, messy reflections, wrinkled packaging, or a weak background.

Edit the provided product image.

After goal:
Create a clean commercial product after image. Remove dust, scratches, clutter, harsh glare, and background distractions. Improve edge definition, label readability, reflections, and shadow quality.

Keep fixed:
Keep the exact product shape, logo, label text, packaging proportions, cap position, color identity, and camera angle. Do not invent new branding or change the product design.

Edit strength:
Medium cleanup. Make it look professionally shot, not artificially rebuilt.

Output:
Square ecommerce-ready product image with crisp focus, realistic studio lighting, and clean negative space.

Quality check:
The product must still match the original item exactly, with readable text and no hallucinated label changes.

If text changes, add: treat all visible label text as locked artwork; preserve it exactly.

5. Layout transformation prompt

Some edits are geometric. In these cases, the prompt should behave like a production instruction, not a style request.

Edit the provided packaging mockup into a flat before/after layout transformation.

After goal:
Convert the 3D packaging view into a clean 2D flattened dieline-style presentation. Show all panels, folds, tabs, and label areas in orthographic view.

Keep fixed:
Preserve the original artwork, colors, logo placement, ingredient text zones, panel order, and package proportions. Do not redesign the brand.

Edit strength:
Structural transformation only.

Output:
White background, precise edges, visible fold lines, no perspective distortion, readable panel layout.

Quality check:
The after image should explain the package structure clearly while preserving the original visual identity.

This same pattern works for turning a room photo into a floor-plan-style concept, a website screenshot into a wireframe, or a poster mockup into a print-ready flat layout.

6. Room makeover prompt

Room edits need constraints around architecture. Otherwise the model may move windows, doors, ceiling height, or furniture scale.

Edit the provided interior photo.

After goal:
Create a before/after room makeover with warmer lighting, cleaner styling, coordinated furniture, less clutter, and a more intentional color palette.

Keep fixed:
Keep the room architecture, window and door positions, ceiling height, floor plan, camera angle, major furniture footprint, and natural light direction.

Allowed changes:
Update decor, textiles, wall color, small furniture, plants, rugs, artwork, cable clutter, and lighting mood.

Edit strength:
Medium makeover. Do not make the room unrecognizable.

Output:
Realistic interior design after image, wide-angle room photo, natural shadows, no warped furniture.

Quality check:
The after image should feel like the same room professionally styled, not a different property.

7. Aesthetic social edit prompt

This is the safer version of the common ChatGPT photo editing prompts for girl search pattern: it improves style without changing identity or body shape.

Edit the provided lifestyle photo.

After goal:
Create an aesthetic social-media-ready edit with cohesive color grading, cleaner background distractions, soft natural light, gentle contrast, and a polished editorial mood.

Keep fixed:
Keep the person's identity, face, body proportions, outfit, pose, expression, and scene logic. Do not reshape the body, change age, or alter facial features.

Style direction:
Choose one: clean minimal, soft film, warm golden hour, cool city night, fresh outdoor lifestyle, or premium editorial.

Edit strength:
Subtle to medium. Preserve realism.

Output:
High-resolution vertical image for social posting, natural skin texture, clean composition.

Quality check:
The after image should look like careful color grading and cleanup, not a synthetic replacement.

Before/after failure fixes

Problem	Add this line to the prompt
The person changes identity	Preserve identity, facial proportions, age, expression, and hairstyle exactly.
The product label changes	Treat all logo and label text as locked artwork; do not rewrite or invent text.
The edit is too subtle	Make the before/after difference clearly visible while keeping the locked elements unchanged.
The edit is too dramatic	Reduce the edit strength; keep the result realistic and close to the source image.
Background looks pasted on	Match the original lighting direction, shadow softness, perspective, and lens depth.
Layout becomes distorted	Use orthographic alignment where relevant; preserve proportions and straight edges.

Quick copy-paste prompt

Use this compact version when you need a fast result:

Edit the provided image into a clear before/after result.

Change:
[Describe the after state.]

Keep unchanged:
[List identity, product details, pose, camera angle, text, layout, architecture, or anything else that must stay fixed.]

Style:
[Realistic / editorial / ecommerce / interior design / social aesthetic / technical layout.]

Strength:
[Subtle / medium / dramatic.]

Quality check:
The final image must clearly show the requested improvement while preserving the locked details from the original image.

FAQ

What should I include in before after image editing prompts?

Include the current before state, the desired after state, locked details, edit strength, output format, and quality check. The locked details are the most important part when editing people, products, rooms, and branded assets.

Can I use these as free photo editor prompts?

Yes. The prompts are written as plain text, so you can copy them into a capable image editing model. For GPT Image 2 examples and prompt patterns, start with the GPT Image 2 prompts library.

What is the best ChatGPT photo editing prompt text?

The best prompt is specific about both change and preservation. Instead of make this photo aesthetic, write what should change, what must stay fixed, and how strong the edit should be.

How do I make before/after edits look realistic?

Keep camera angle, lighting direction, shadows, identity, and proportions consistent. Ask for a medium-strength edit first, then increase the change only if the result is too subtle.

Should I ask for a split-screen before/after image?

Use a split-screen when the comparison itself is the deliverable. If you only need the final edited asset, ask for the after image and include a quality check that compares it against the original.